KR100668300B1 - Bitrate scalable speech coding and decoding apparatus and method thereof - Google Patents

Abstract

The present invention relates to an SNR bit rate extended speech encoding and decoding apparatus and a method thereof, wherein the encoding apparatus includes a base layer, a sound quality enhancement layer, and a multiplexer, and the base layer includes an input speech signal by linear predictive encoding. Filter, generate an excitation signal corresponding to the speech signal filtered by the fixed codebook search and the adaptive codebook search, and the sound quality enhancement layer searches for the fixed codebook using the parameters generated according to the fixed codebook search in the base layer; Fixed codebook search of base layer The fixed codebook is searched by using the signal that removes the contribution of the fixed codebook of the base layer from the target signal and the signal obtained by synthesizing the previous fixed codebook in the sound quality enhancement layer as the target signal of the sound quality enhancement layer. Enhances the signal generated by the base layer and at least one sound quality By multiplexing the signal generated in the layer, the encoding apparatus according to the present invention is compatible with the existing standardized speech codec, can reduce the amount of calculation, and can provide better sound quality.

Description

Bitrate scalable speech coding and decoding apparatus and method

1 is a block diagram of a bit rate extended speech encoding apparatus according to an exemplary embodiment of the present invention.

FIG. 2 is an exemplary diagram of positions of pulses searched by the base layer fixed codebook search unit shown in FIG. 1 and positions of pulses searched by the sound quality enhancement layer fixed codebook search unit.

3 is a block diagram of a bit rate extended speech decoding apparatus according to an exemplary embodiment of the present invention.

4 is an operation flowchart of a bit rate extended speech encoding method according to an embodiment of the present invention.

5 is a flowchart illustrating an operation of a bit rate extended speech decoding method according to an exemplary embodiment of the present invention.

6 is a block diagram of a bit rate extended speech encoding apparatus according to another exemplary embodiment of the present invention.

FIG. 7 is a block diagram illustrating an exemplary embodiment of a gain value difference quantizer of the sound quality enhancement layer illustrated in FIG. 6.

8 is a block diagram of a bit rate extended speech decoding apparatus according to another exemplary embodiment of the present invention.

9 is a diagram illustrating a position of a pulse searched by a base layer fixed codebook search and a position of a pulse searched by a sound quality enhancement layer fixed codebook search in the bit rate extension speech decoding apparatus of FIG. 8.

10 is a flowchart illustrating an operation of a bit rate extended speech encoding method according to another exemplary embodiment of the present invention.

11 is an operation flowchart of a bit rate extended speech decoding method according to another preferred embodiment of the present invention.

The present invention relates to a speech codec (codec) using a Kelp (Code Excited Linear Prediction, hereinafter referred to as CELP) algorithm, and in particular, speech coding and decoding for extending the SNR (Signal to Noise Ratio) bit rate to improve sound quality. It relates to an apparatus and a method thereof.

The speech codec having the CELP structure is the most widely used in the current mobile communication system, and is based on linear prediction coding (hereinafter, referred to as LPC). The voice codec having the CELP structure has a different data rate and bandwidth depending on the type of service.

However, in the general voice codec, since the bit rate and bandwidth are set in the encoding device, the bit rate and bandwidth cannot be selected in the decoding device. In addition, when multicasting is performed to transmit packet information from one transmitter to multiple receivers on the network, if the voice codec of the transmitter has a fixed bit rate, the packet information transmitted to the receivers requiring different bit rates may be used. The quality may deteriorate.

In order to improve this, a speech codec employing a bit rate extended speech coding scheme has been proposed. This voice codec constitutes a bit stream so that not only the information of the base codec but also the information to more accurately correct the signal to be restored are added.

Conventional bit rate extension speech coding schemes can be broadly classified into signal to noise ratio (SNR) bit rate extension methods and bandwidth extension methods.

Speech encoding by the SNR bit rate extension method encodes and decodes a speech signal using a hierarchical coding scheme. That is, the speech signal is encoded by dividing the speech signal into a base layer and a speech enhancement layer. The base layer transmits only information that can restore the minimum sound quality. The sound quality enhancement layer transmits additional information to improve sound quality.

However, the proposed SNR bit rate extended speech encoding apparatus is configured to independently encode the base layer and the sound quality enhancement layer. Therefore, since the calculation of correlation and energy between the target signal (or target vector) and the impulse response required when searching the fixed codebook is performed in the base layer and the sound quality enhancement layer, the parameters for the fixed codebook search are obtained. A lot of computation is required.

In addition, the proposed SNR bit rate extended speech coder has a disadvantage in that it is not compatible with the existing standardized CELP speech coder by changing the structure of the existing standardized CELP speech coder to further operate the sound quality enhancement layer. .

The present invention provides a speech encoding and decoding apparatus and method for extending an SNR bit rate compatible with existing standardized speech codecs, including a fixed codebook of a conventional standardized speech codec and a fixed codebook having a multi-layered structure. To provide.

Another object of the present invention is to provide an SNR bit rate extended speech encoding and decoding apparatus and method for reducing the amount of computation for obtaining parameters for fixed codebook search.

Another technical problem to be solved by the present invention is SNR for searching a fixed codebook of a sound quality enhancement layer by using a target signal from which contribution of the fixed codebook found in the base layer and a synthesized excitation signal of the sound quality enhancement layer are removed. Disclosed are a bit rate extended speech encoding and decoding apparatus and a method thereof.

Another technical problem to be solved by the present invention is to allow overlapping of the position of the pulse searched in the base layer and the position of the pulse searched in the sound quality enhancement layer, thereby overcoming the limitations of the algebraic codebook. The present invention provides a decoding apparatus and a method thereof.                         

Another object of the present invention is to provide an apparatus and method for SNR bit rate extension speech encoding and decoding capable of reducing quantization bits for gain values of fixed codebooks in a sound quality enhancement layer.

In order to achieve the above technical problem, the present invention provides a communication system comprising: a base layer for filtering an input speech signal using linear predictive coding and generating an excitation signal of the filtered speech signal by fixed codebook searching and adaptive codebook searching; And at least one sound quality enhancement layer for searching a fixed codebook using a parameter obtained by the fixed codebook search in the base layer, and multiplexing a signal generated in the base layer and a signal generated in the sound quality enhancement layer. And a multiplexer for outputting the multiplexed signal.

According to an aspect of the present invention, there is provided an apparatus including: a base layer configured to linearly predictively encode an input speech signal and generate an excitation signal corresponding to the filtered speech signal by fixed codebook searching and adaptive codebook searching; And a fixed codebook search unit for searching a fixed codebook using a parameter generated according to the fixed codebook search in the base layer, a first fixed codebook gain value generated by the fixed codebook search in the base layer, and the fixed codebook. A plurality of sound quality enhancement layers including a gain difference quantizer for detecting a difference between the second fixed codebook gain values output from the searcher and quantizing the detected difference, and a signal generated in the base layer and the sound quality enhancement layer Provided is a speech signal encoding apparatus including a multiplexer for multiplexing a signal generated by a.

In order to achieve the above technical problem, the present invention provides a speech signal decoding apparatus for decoding a speech signal encoded by being divided into a base layer and at least one sound quality enhancement layer, and decodes the encoding information of the encoded speech signal in the base layer. A first decoding unit for performing; A second decoding unit for restoring encoding information in a sound quality enhancement layer among the encoded speech signals according to an operating environment of the speech signal decoding apparatus; A calculation unit configured to calculate a signal restored in the first decoding unit and a signal restored in the second decoding unit according to an operating environment of the speech signal decoding apparatus; Provided is a speech signal decoding apparatus comprising a speech signal recovery unit for recovering a speech signal by synthesizing a signal output from the calculation unit using the linear prediction coding coefficients output from the first decoding unit.

The first decoding unit includes: a linear prediction coding coefficient decoder for decoding linear prediction coding coefficient quantization information included in the encoding information in the base layer; A first fixed codebook decoder which decodes a fixed codebook index included in the encoding information in the base layer; An adaptive codebook decoder which decodes an adaptive codebook index included in the encoding information in the base layer; Preferably, a gain value decoding unit for decoding the fixed codebook gain value and the adaptive codebook gain value included in the encoding information in the base layer, respectively.

The second decoding unit includes: a gain value difference decoder which decodes quantization information of a difference between fixed codebook gain values included in encoded information in the speech enhancement layer; Preferably, a second fixed codebook decoder for decoding the fixed codebook index included in the encoding information in the sound quality enhancement layer is included.

The second decoding unit includes: a gain value difference decoding unit for decoding quantization information of the difference between the fixed codebook logscale gain values included in the encoding information in the sound quality enhancement layer; Preferably, a second fixed codebook decoder for decoding the fixed codebook index included in the encoding information in the sound quality enhancement layer is included.

The second decoding unit includes: a gain value difference decoding unit for decoding the quantization information of the difference between the fixed codebook logscale gain values included in the encoding information in the speech enhancement layer; Preferably, the fixed codebook decoder includes a fixed codebook index included in the encoded information in the sound quality enhancement layer.

According to an aspect of the present invention, there is provided a basic layer processing step of extracting a linear predictive coding coefficient of an input speech signal and generating an excitation signal corresponding to the input speech signal by a fixed codebook search and an adaptive codebook search. ; A sound quality enhancement layer processing step of searching for a fixed codebook using a parameter generated according to the fixed codebook search in the base layer processing step; It provides a speech signal encoding method comprising the step of multiplexing the signal generated by the base layer processing step and the sound quality enhancement layer processing step.

The sound quality enhancement layer processing step is preferably performed in a plurality of steps.

According to an aspect of the present invention, there is provided a speech signal decoding method for decoding a speech signal encoded by a base layer and at least one sound quality enhancement layer, the method comprising: decoding the encoded speech signal; Selectively transmitting the codebook for the base layer decoded in the decoding step and the codebook for the sound quality enhancement layer according to the operating condition of the speech signal decoding; And generating a reconstructed speech signal by combining the selectively transmitted codebook and the linear prediction coefficients decoded in the decoding step.

In order to achieve the above technical problem, the present invention includes a base layer for filtering an input speech signal using linear predictive coding, and generating an excitation signal of the filtered speech signal by fixed codebook search and adaptive codebook search; And a sound quality enhancement layer for searching for a fixed codebook by using a signal from which the contribution of the fixed codebook of the base layer is removed from the fixed codebook search target signal of the base layer, and the signal generated in the base layer and the sound quality. The present invention provides a speech signal encoding apparatus including a multiplexer for multiplexing a signal generated in an enhancement layer and outputting the multiplexed signal.

The fixed codebook contribution y ₂ (n) of the base layer is calculated based on the following equation using the impulse response h (n) of the synthesis filter and the fixed codebook c _G multiplied by the quantization gain value of the fixed codebook of the base layer. desirable.

The sound quality enhancement layer may further remove a signal obtained by synthesizing the fixed codebook signal generated in the sound quality enhancement layer by using the linear prediction coding coefficients from the target signal of the base layer.

In the fixed codebook search of the sound quality enhancement layer, between the log scale value of the first gain value obtained by the fixed codebook search of the base layer and the log scale value of the second gain value obtained by the fixed codebook search in the sound quality enhancement layer. It is preferable to further include a function of obtaining a quantized gain value of the sound quality enhancement layer by using the result of quantizing the difference and multiplying the quantized gain value by the fixed codebook vector obtained by the fixed codebook search in the sound quality enhancement layer.

The sound quality enhancement layer may perform the fixed codebook search after the cognitive weighting filtering of the target signal.

According to an aspect of the present invention, there is provided an apparatus, including: a base layer configured to linearly predictively code an input speech signal and generate an excitation signal corresponding to the filtered speech signal by fixed codebook searching and adaptive codebook searching; A searcher for searching for a fixed codebook by using a signal from which the fixed codebook contribution of the base layer is removed from the fixed codebook search target signal of the base layer as a fixed codebook search target signal of a sound quality enhancement layer, by the fixed codebook search of the base layer And a log scale gain value difference quantizer for detecting a difference between the generated log scale gain value of the first fixed codebook and the log scale gain value of the second fixed codebook output from the fixed codebook search unit, and quantizing the detected difference. And a multiplexer for multiplexing a signal generated in the base layer and a signal generated in the sound quality enhancement layer, wherein the sound quality enhancement layer is fixed using a linear prediction coding coefficient in the sound quality enhancement layer. Codebook synthesized signal of the sound quality enhancement layer And it provides an audio signal encoding apparatus according to claim 1, further removed from the constant codebook search target signal.

According to an aspect of the present invention, there is provided a method, comprising: a base layer processing step of extracting a linear prediction coefficient of an input speech signal and generating an excitation signal corresponding to the input speech signal by a fixed codebook search and an adaptive codebook search; A sound quality enhancement layer processing step of searching for a fixed codebook by using a signal from which the fixed codebook contribution of the base layer is removed from the fixed codebook search target signal of the base layer as a fixed codebook search target signal of a sound quality enhancement layer; It provides a speech signal encoding method comprising the step of multiplexing the signal generated by the base layer processing step and the sound quality enhancement layer processing step.

Hereinafter, a bit rate extended speech encoding and decoding apparatus and a method thereof according to an embodiment of the present invention will be described.

1 is a functional block diagram of a bit rate extended speech encoding apparatus according to an embodiment of the present invention. Referring to FIG. 1, a speech encoding apparatus according to an embodiment of the present invention has a multi-layer fixed codebook structure including a base layer 100 and a sound quality enhancement layer 130.

In the base layer 100, encoding information for restoring the minimum sound quality is generated. The base layer 100 is similar to the configuration of the existing standardized CELP speech coder. Accordingly, the base layer 100 filters the input speech signal by linear predictive coding to generate an excitation signal corresponding to the input speech signal.

The base layer 100 includes a preprocessing unit 102, an LPC coefficient extraction and vector quantizer 104, a synthesis filter 106, a subtractor 108, a perceptual weighting filter 110, and a pitch. Analysis unit 112, pitch contribution removal unit 115, fixed codebook search unit 117, fixed codebook 119, first multiplier 121, adder 123, adaptive codebook 124, 2 multiplier 126, a gain value quantizer 129.

The preprocessing unit 102 removes the DC component from the audio signal input through the line 101. That is, the preprocessing unit 102 filters the input voice signal using a high pass filter to remove noise components of the low frequency band of the input voice signal. The high pass filter H _h1 (n) used has a transfer function as shown in equation (1).

The signal output from the preprocessing unit 102 is transmitted to the LPC coefficient extraction and vector quantizer 104 via line 103.

The LPC coefficient extraction and vector quantizer 104 extracts the LPC coefficients of the signal output from the preprocessing unit 102. The extracted LPC coefficients are vector quantized by the LPC coefficient extraction and vector quantizer 104. Vector quantization information of the LPC coefficients is sent via line 105 to synthesis filter 106 and multiplexer 140.

A synthesis filter 106 outputs a synthesized signal corresponding to an excitation signal input through the line 128 using the vector quantization information of the LPC coefficients. The synthesized signal is output to subtractor 108 via line 107.

The subtractor 108 subtracts the synthesized signal input through the line 107 from the signal output from the preprocessing unit 102 input through the line 103 to generate a difference signal. The difference signal is transmitted via the line 109 to the cognitive weighting filter 110.

The cognitive weighting filter 110 allows the quantization noise to be less than or equal to the masking threshold in order to use the masking effect of the human auditory structure. Accordingly, the cognitive weighting filter 110 outputs a signal including a weight to the pitch analyzer 112 so that the quantization noise of the difference signal is minimized.

The pitch analyzer 112 searches for an open-loop pitch and a close-loop pitch with respect to the signal output from the cognitive weight filter 110. That is, the pitch analyzer 112 divides the signal output from the cognitive weighting filter 110 into a plurality of subframes, analyzes the pitch of each subframe, and outputs an index and a gain value of the adaptive codebook. The index of the adaptive codebook is transmitted to the multiplexer 140 through the line 114 while being transmitted to the pitch contribution remover 115 and the adaptive codebook 124 through the line 113. The gain value of the adaptive codebook is also provided to a gain value quantizer 129.

The pitch contribution remover 115 detects a target signal (or target vector) required for the fixed codebook search from the output signal of the cognitive weight filter 110 based on the index of the adaptive codebook. The pitch contribution remover 115 subtracts the pitch contribution y ₁ (n) from the line 111 to transmit the fixed codebook search target signal to the fixed codebook search unit 117 of the base layer 100 through the line 116. The fixed codebook search unit 131 of the sound quality enhancement layer 130 is output. The pitch contribution y ₁ (n) is obtained by equation (2).

In Equation 2, AC _G (n) is a value multiplied by an adaptive codebook gain value.

The fixed codebook search unit 117 calculates the correlation d (n) between the target signal and the impulse response h (n) using the target signal x '(n) input through the line 111.

For example, assuming that the size of the subframe is 40 samples and the number of pulses in each layer is four, the correlation d (n) may be defined as shown in Equation 3 below.

In Equation 3, h (i-n) is an impulse response, and x '(n) is a target signal.

The impulse response h (n) and the correlation d (n) are provided to the fixed codebook search unit 131 of the sound quality enhancement layer 130 through the line 118 '.

The fixed codebook search unit 117 searches for a fixed codebook in the form of an algebraic codebook configured as shown in Table 1 based on the impulse response h (n) and the correlation d (n).

Referring to Table 1, in the fixed codebook search unit 117, the magnitude of the pulse is not zero only in four positions. Therefore, the correlation C, which is the sum of the magnitudes of the correlation d (n) of each pulse using the symbol s of the pulse and the correlation d (n), may be defined as in Equation 2. The fixed codebook search unit 117 detects the correlation C by equation (4).

In Equation 4, m _i represents the position of the i-th pulse, s _i represents the sign of the i-th pulse. The fixed codebook detector 117 detects the energy E of the impulse response h (n) of the synthesis filter 106 by the equation (5).

?? (m _i , m _j ) in Equation 5 is a correlation between the position of the i-th pulse and the impulse response signal h (n) with respect to the position of the j-th pulse, s _i is the sign of the i-th pulse, and s _j is the sign of the j th pulse.

The fixed codebook search unit 117 stores the correlation C and the energy E of the impulse response h (n). The correlation degree C is stored by dividing the sign [d (i)] and its absolute value. sign [d (i)] is the sign of d (i). The energy E is stored in the form as shown in Equation 6.

Equation 5 for energy E may be redefined as in Equation 7.

The fixed codebook search unit 117 provides the detected correlation C and energy E to the fixed codebook search unit 131 of the sound quality enhancement layer 130 through the line 118 ", while detecting the detected correlation C and energy E. The fixed codebook consisting of algebraic codebooks is searched using E. When the fixed codebook index and the gain value are obtained by the fixed codebook search, the fixed codebook search unit 117 multiplies the fixed codebook index with the fixed codebook 119. And transmits the gain value to the gain value quantizer 129.

The fixed codebook 119 outputs a fixed codebook vector of the base layer 100 based on the index input through the line 118. The fixed codebook vector output from the fixed codebook 119 is provided to the first multiplier 121 via the line 120.

The first multiplier 121 multiplies the pulse position and the sign information by the quantization gain value G _c for the gain value of the fixed codebook provided by the gain value quantizer 129 and outputs the result through the line 122. do. The signal output over line 122 is a vector of fixed codebooks. The quantization gain value G _c is provided from a gain value quantizer 129.

When the adaptive codebook index is applied through the line 113, the adaptive codebook 124 outputs position information and sign information of a pulse corresponding to the adaptive codebook index. The adaptive codebook vector output over line 125 is provided to a second multiplier 126.

The second multiplier 126 multiplies the quantized gain value G _p for the gain value of the adaptive codebook by the adaptive codebook vector transmitted via the line 125 and outputs the result via line 127. The signal output through the line 127 is a vector of the adaptive codebook multiplied by the gain value G _p . The quantized gain value G _p is provided from a gain value quantizer 129.

The adder 123 adds the fixed codebook vector multiplied by the gain value G _c inputted through the line 122 and the adaptive codebook vector multiplied by the gain value G _p inputted through the line 127 to obtain an excitation signal. The excitation signal is output to the synthesis filter 106 via line 128.

The gain value quantizer 129 quantizes the gain value of the fixed codebook output from the fixed codebook search unit 117 and the gain value of the adaptive codebook output from the pitch analyzer 112, respectively. The gain value G _c obtained by quantizing the gain value of the fixed codebook is output to the first multiplier 121, and the gain value G _p obtained by quantizing the gain value of the adaptive codebook is output to the second multiplier 126. The quantized gain value G _c is also provided to the gain value quantizer 134 included in the sound quality enhancement layer 130.

The sound quality enhancement layer 130 is to provide additional bits in addition to the bits provided in the base layer 100 in order to improve the restored sound quality. For example, when the base layer 100 provides a bit rate of 8 kbps, the sound quality enhancement layer 130 may provide an additional bit rate of 4 kbps. 1 illustrates a configuration in which one voice enhancement layer 130 is connected to the base layer 100 for convenience of description, but a plurality of voice enhancement layers may be connected to the base layer 100.

The sound quality enhancement layer 130 includes a fixed codebook search unit 131 and a gain value quantizer 134. The fixed codebook search unit 131 has an impulse response signal h (n) provided through the line 118 ', d (n) which is a correlation between the target signal and the impulse response signal h (n), and the sign of the pulse. A fixed codebook consisting of logarithmic codebooks is searched using the correlation C corresponding to the magnitude information of d (n) detected using the diagram d (n) and the energy E of the impulse response signal h (n).

As described above, the fixed codebook search unit 131 performs a fixed codebook search on the same target signal as the target signal searched by the fixed codebook search unit 117. The fixed codebook search unit 131 uses an algebraic codebook. The fixed codebook search unit 131 minimizes MSE (Mean Square Error) of the target signal (target vector) and finds a vector c _k that maximizes Equation 6. The found vector c _k becomes a fixed codebook vector.

In Equation 8 Φ represents the correlation between the impulse response h (n). The d (n) and Φ use values provided by the base layer 100. Φ is provided from the fixed codebook search unit 117. Accordingly, the fixed codebook search unit 131 can reduce the amount of computation required when searching for the fixed codebook.

Assuming that the degree of the fixed codebook vector of the base layer 100 is 40 and that the base layer 100 and the sound quality enhancement layer 130 find four non-zero pulses, respectively, the base layer 100 is fixed. Since the codebook 117 first finds four pulses and the fixed codebook search unit 131 of the sound quality enhancement layer 130 finds four pulses, the fixed codebook search unit 131 finds four pulses found in the base layer 100. Consider the effects of the pulses. Therefore, the correlation C 'obtained from the fixed codebook search unit 131 may be defined as in Equation 9, and the energy E' may be defined as in Equation 10.

Equation 9 may be redefined as in Equation 11 using the correlation C value defined in Equation 4.

The fixed codebook search unit 131 may detect the energy E by a redefined operation as shown in Equation 12 in order to reduce the complexity of the search process.

Equation 12 may be redefined as in Equation 13 using the energy E defined in Equation 7.

The correlation C 'and the energy E' may be stored before the fixed codebook search in the sound quality enhancement layer 130 to simplify the fixed codebook search process.

Of the fixed codebook search unit 131 for obtaining the sign information and the position information of the pulse of the sound quality enhancement layer 130 using the above-described correlations C 'and energy E'. The process is the same as that performed by the fixed codebook search unit 117 of the base layer 100. In this case, the position information of the pulse searched in the base layer 100 and the position information of the pulse searched in the sound quality enhancement layer may be the same.

FIG. 2 is a diagram for describing a position of a pulse searched by the fixed codebook search unit 117 and a position of a pulse searched by the fixed codebook search unit 131 in the bit rate extended speech encoding apparatus of FIG. 1.

Referring to FIG. 2, the position of the pulse searched in the fixed codebook search 201 may be the same as the position of the pulse searched in the sound quality enhancement layer fixed codebook search 202. Accordingly, the magnitude of the pulse of the final fixed codebook has multiple magnitudes including the magnitude of the fixed codebook pulses of the base layer 100 and the sound quality enhancement layer 130. Thus, the magnitude of the pulses of algebraic codebooks does not have only +1 or -1.

The fixed codebook search unit 131 provides the fixed codebook vector obtained according to the search result to the multiplexer 140 and provides the gain value of the fixed codebook to the gain value quantizer 134. The fixed codebook index in the sound quality enhancement layer 130 may be composed of pulse code information and pulse position information.

As such, the fixed codebook index found in the sound quality enhancement layer 130 is not stored for the next frame and thus does not affect the operation of the base layer 100.

The gain difference quantizer 134 obtains a difference between the gain value 132 of the fixed codebook obtained by the fixed codebook search unit 131 and the gain value G _c of the fixed codebook quantized in the base layer 100, Quantize the car. Accordingly, since the gain difference quantization information G _diff is transmitted from the gain difference quantizer 134 to the multiplexer 140 through the line 135, the sound quality enhancement layer 130 is applied to the gain value of the fixed codebook. Quantization bits can be reduced.

The multiplexer 140 may include the LPC coefficient quantization information, the fixed codebook index, the adaptive codebook index, the gain value quantization information provided from the base layer 100, and the fixed codebook index of the sound quality enhancement layer provided from the sound quality enhancement layer 130. The value difference quantization information is output as a bit stream.

The bit streams of the base layer 100 and the sound quality enhancement layer 130 are separately transmitted. That is, as shown in FIG. 1, the bit streams of the sound quality enhancement layer 130 are transmitted after the bit streams of the base layer 100. do. Accordingly, the bit stream can be easily separated at the bit rate required for the decoding apparatus according to the network traffic conditions. For example, when the channel characteristic of the decoding apparatus is poor and only the bit stream of the base layer can be received, the decoding apparatus receives only the bit stream of the base layer from the bit stream transmitted by the bit rate extension speech encoding apparatus of FIG. 1. can do.

3 is a block diagram of a bit rate extended speech decoding apparatus according to an exemplary embodiment of the present invention.

Referring to FIG. 3, the apparatus for decoding a bit rate extension speech includes a demultiplexer 301, an LPC coefficient decoder 302, a gain value decoder 303, a first fixed codebook decoder 304, and an adaptive codebook decoder. 305, gain difference decoder 306, second fixed codebook decoder 307, first adder 308, second adder 309, first select switch 310, second select switch 311 ), A first multiplier 312, a second multiplier 313, a third adder 314, a synthesis filter 315, and a post processor 316.

The bit rate extended speech decoding apparatus may selectively receive a bit stream transmitted from the bit rate extended speech encoding apparatus. That is, if only the bit stream for the base layer is received in the bit stream, the sound quality of the base layer can be restored, and if both the bit streams for the base layer and the sound quality enhancement layer are received, a more improved sound quality can be provided.

The demultiplexer 301 demultiplexes the received bit stream with information of each module and outputs the demultiplexer. That is, the demultiplexer 301 converts the LPC coefficient quantization information into the LPC coefficient decoder 302, the gain value quantization information into the gain value decoder 303, and the gain difference quantization information into the gain value difference decoder 306. The fixed codebook index of the sound quality enhancement layer is provided to the second fixed codebook decoder 307, the fixed codebook index to the first fixed codebook decoder 304, and the adaptive codebook index to the adaptive codebook decoder 305, respectively. do.

The structure of the LPC coefficient decoder 302 is determined by the LPC coefficient extraction and vector quantizer 104 on the encoding apparatus side, and restores the LPC coefficients from the input LPC coefficient quantization information. The restored LPC coefficients are provided to the synthesis filter 315 and the post processor 316.

The structure of the gain value decoder 303 is determined by the gain value quantizer 129 on the encoding device side. The gain value decoder 303 decodes the input gain value quantization information. The gain value quantization information includes an adaptive codebook gain value and a fixed codebook gain value. Therefore, the adaptive codebook gain value g _p and the fixed codebook gain value g _c in the base layer 100 are output from the gain value decoding unit 303, respectively.

The first fixed codebook decoder 304 decodes the fixed codebook index of the input base layer 100 and outputs the fixed codebook of the base layer 100. The fixed codebook decoding method is determined by the search method in the fixed codebook search unit 117 of the encoding apparatus. The adaptive codebook decoder 305 decodes the input adaptive codebook index and outputs the adaptive codebook of the base layer 100.

The LPC coefficient decoder 302, the gain value decoder 303, the first fixed codebook decoder 304, and the adaptive codebook decoder 305 described above are the base layer 100 transmitted from the demultiplexer 301. It may be defined as a first decoding unit for decoding the encoding information in the.

The operation of the gain value difference decoder 306 and the second fixed codebook decoder 307 depends on the network traffic state or the processing capacity of the receiving terminal.

If it is determined that the gain value difference decoder 306 and the second fixed codebook decoder 307 are operated, the gain value difference decoder 306 decodes the input gain value difference quantization information. The second fixed codebook decoder 307 decodes the fixed codebook index of the input sound quality enhancement layer. The gain value difference decoding method is determined by the gain value difference quantizer 134 on the encoding apparatus side. The decoding method of the second fixed codebook decoding unit 307 is determined by the second fixed codebook searching unit 131 on the encoding apparatus side.

The gain value difference decoder 306 and the second fixed codebook decoder 307 may be regarded as a second decoding unit for decoding the encoding information in the sound quality enhancement layer 130 transmitted from the demultiplexer 301.

The first adder 308 adds the gain value g _c of the decoded fixed codebook output from the gain value decoding unit 303 and the decoded gain value difference g _diff output from the gain value difference decoding unit 306. The output of the first adder 308 is the gain value of the sound quality enhancement layer upon decoding.

The second adder 309 selects the fixed codebook of the sound quality enhancement layer 130 decoded by the second fixed codebook decoder 307 and the fixed codebook of the base layer 100 decoded by the first fixed codebook decoder 304. Add. Therefore, the signal output from the second adder 309 may be defined as shown in Equation (13).

In Equation 14, c (n) is a fixed codebook in the base layer, and c '(n) is a fixed codebook in the sound quality enhancement layer.

Accordingly, the fixed codebook pulse in the decoding apparatus has an algebraic codebook pulse structure having multiple magnitudes by accumulating algebraic codebooks of the base layer and the sound quality enhancement layer. Accumulating the algebraic codebooks is to compensate for the disadvantages of the existing fixed codebook structure in which all pulses have the same magnitude. Therefore, the accumulated pulse code of the algebraic codebook has a code suitable for the target signal.

The first selection switch 310 selectively transmits the fixed codebook gain value g _c decoded by the gain value decoder 303 and the signal output from the first adder 308. That is, when the decoding apparatus operates as the base layer, the first selection switch 310 transmits the fixed codebook gain value g _c output from the gain value decoding unit 303, and when the corresponding decoding apparatus operates as the sound quality enhancement layer. The first selection switch 310 transmits a gain value output from the adder 308.

The second selection switch 311 selectively transmits the signal output from the second adder 309 and the fixed codebook in the base layer 100 output from the first fixed codebook decoder 304. That is, when the decoding apparatus is not operated in the sound quality enhancement layer, the second selection switch 311 transmits a signal output from the first fixed codebook decoder 304 and the decoding apparatus operates in the sound quality enhancement layer. In this case, the second selection switch 311 transmits a signal output from the second adder 309.

The first multiplier 312 multiplies and outputs a gain value output from the first selection switch 310 to the fixed codebook output from the second selection switch 311.

The second multiplier 313 multiplies the decoded adaptive codebook output from the adaptive codebook decoder 305 by the gain value g _p of the adaptive codebook output from the gain value decoder 303.

The third adder 314 adds the information about the fixed codebook output from the first multiplier 312 and the information about the adaptive codebook output from the second multiplier 313 to generate a restored excitation signal.

The first adder 308, the second adder 309, the third adder 314, the first multiplier 312, the second multiplier 313, the first selector switch 310, and the second selector switch ( 311 may be defined as a calculation unit that calculates the signals decoded by the first decoding unit and the second decoding unit according to an operating environment of the decoding apparatus.

The synthesis filter 315 synthesizes the excitation signal provided from the adder 314 using the reconstructed LPC coefficients provided from the LPC coefficient decoder 302 to restore the speech signal.

The post processor 316 serves to improve the sound quality of the voice signal transmitted from the synthesis filter 315. That is, the post processor 316 is a high pass filter for filtering the signal output from the synthesis filter 315 using the LPC coefficients provided from the LPC coefficient decoder 302 in order to improve the sound quality of the voice signal. Pass Filtering).

The synthesis filter 315 and the post processor 316 described above may be defined as a reconstruction unit for reconstructing a voice signal by combining the signal output from the operation unit with the LPC coefficient output from the LPC coefficient decoder 302.

4 is a flowchart illustrating an operation of a bit rate extended speech encoding method according to an embodiment of the present invention.

In operation 401, the apparatus for encoding a speech signal preprocesses an input speech signal as in the preprocessing unit 102 of FIG. 1. In operation 402, the apparatus for encoding a speech signal extracts LPC coefficients from a preprocessed speech signal and generates quantization information of the extracted LPC coefficients.

In operation 403, the apparatus for encoding a speech signal synthesizes the excitation signal as in the synthesis filter 106 of FIG. 1 using the generated quantization information of the LPC coefficients. In operation 404, the apparatus for encoding a speech signal detects an LPC residual signal by subtracting the synthesized signal from the preprocessed signal. In operation 405, the apparatus for encoding a speech signal filters the detected LPC residual signal as in the cognitive weighting filter 110 of FIG. 1 and outputs a cognitive weighted signal.

In operation 406, the apparatus for encoding a speech signal analyzes the pitch of the perceptually weighted signal as in the pitch analyzer 112 of FIG. 1 to obtain an index and a gain value of the adaptive codebook. Like the pitch contribution remover 115 of FIG. 1, the pitch contribution is removed from the cognitively weighted signal based on the index of the adaptive codebook to detect the target signal required for the fixed codebook search.

In operation 407, the apparatus for encoding a speech signal searches for a base layer fixed codebook as in the first fixed codebook search unit 117 of FIG. 1 to generate a fixed codebook gain value and a fixed codebook index. In operation 408, the apparatus for encoding a speech signal quantizes the detected fixed codebook gain value and the detected adaptive codebook gain value as in the gain value quantizer 129 of FIG.

In operation 409, the apparatus for encoding a speech signal searches for a sound quality enhancement layer fixed codebook using parameters such as correlations C and d (n) and energy E in the base layer. The gain of the sound quality enhancement layer fixed codebook and the index of the sound quality enhancement layer fixed codebook are generated by searching the sound quality enhancement layer fixed codebook.

In operation 410, the apparatus for encoding a speech signal quantizes a difference between a gain value of the base layer fixed codebook and a gain value of the fixed codebook of the sound quality enhancement layer. The fixed codebook search and gain value quantization processes in the sound quality enhancement layer may be divided into a plurality of processes as described with reference to FIG. 1. When the processing of the sound quality enhancement layer is divided into a plurality of steps, the quality of the speech signal restored by that can be improved.

In operation 411, the apparatus for encoding a speech signal may include LPC coefficient quantization information obtained through the above steps, a fixed codebook index of a base layer, an adaptive codebook index of a base layer, a gain value of a fixed codebook of a base layer, and a gain of an adaptive codebook of a base layer. A value, a fixed codebook index of the sound quality enhancement layer, and the gain value quantization information are multiplexed in a bit stream to be transmitted to the speech signal decoding apparatus.

5 is a flowchart illustrating an operation of a bit rate extended speech decoding method according to an exemplary embodiment of the present invention.

In operation 501, the apparatus for decoding a speech signal demultiplexes a received bit stream with information of each component, such as the demultiplexer 301 of FIG. 3.

In operation 502, the voice signal decoding apparatus decodes the demultiplexed signal. That is, the LPC coefficient decoder 302, the gain value decoder 303, the first fixed codebook decoder 304, the adaptive codebook decoder 305, the gain value difference decoder 306, and the second of FIG. Like the fixed codebook decoder 307, the demultiplexed signal is decoded.

In operation 503, the speech signal decoding apparatus restores the sound quality enhancement layer fixed codebook gain value by a predetermined operation. The speech signal decoding apparatus reconstructs the fixed codebook gain value of the sound quality enhancement layer by adding a difference between the decoded fixed codebook gain value and the received gain value as quantization information of the fixed codebook gain value of the sound quality enhancement layer.

In operation 504, the voice signal decoding apparatus selectively transmits the fixed codebook of the sound quality enhancement layer and the fixed codebook of the base layer according to the operating conditions of the voice signal decoding apparatus, and optionally the gain value. That is, when the audio signal decoding apparatus is operated in the sound quality enhancement layer, the fixed codebook of the sound quality enhancement layer is multiplied by the gain value of the fixed codebook of the reconstructed sound quality enhancement layer. On the other hand, if the speech signal encoding apparatus does not operate in the sound quality enhancement layer, the fixed codebook multiplying the gain value of the fixed codebook of the base layer is transmitted to the decoded fixed codebook of the base layer.

In operation 505, the apparatus for decoding speech signals synthesizes the fixed codebook selectively transmitted in operation 504 using the LPC coefficients decoded in operation 502.

In operation 506, the voice signal decoding apparatus post-processes the post-processing unit 316 to generate a reconstructed voice signal.

6 is a functional block diagram of a bit rate extended speech encoding apparatus according to another exemplary embodiment of the present invention. Referring to FIG. 6, the apparatus for encoding a bit rate extension speech codec has a multilayer fixed codebook structure including a base layer 600 and a sound quality enhancement layer 630.

In the base layer 600, encoding information for restoring the minimum sound quality is generated. The base layer 600 is similar to the configuration of the existing standardized CELP speech coder. Accordingly, the base layer 600 filters the input speech signal by linear predictive coding and generates an excitation signal corresponding to the filtered speech signal. The excitation signal is generated by fixed codebook search and adaptive codebook search.

Base layer 600 includes preprocessing unit 602, LPC coefficient extraction and vector quantizer 604, synthesis filter 606, subtractor 608, perceptual weighting filter 610, pitch Analysis unit 612, pitch contribution removal unit 615, fixed codebook search unit 617, fixed codebook 619, first multiplier 621, adder 623, adaptive codebook 624, Two multipliers 626 and a gain value quantizer 629.

The preprocessing unit 602 removes the DC component from the voice signal input through the line 601. That is, the preprocessing unit 602 filters the input voice signal using a high pass filter to remove noise components of the low frequency band of the input voice signal. The high pass filter used is the same as the high pass filter of the preprocessing unit 102 of the base layer 100 in one embodiment of the invention. The signal output from preprocessing unit 602 is sent via line 603 to LPC coefficient extraction and vector quantizer 604.

The LPC coefficient extraction and vector quantizer 604 extracts the LPC coefficients of the signal output from the preprocessing unit 602. The extracted LPC coefficients are vector quantized by LPC coefficient extraction and vector quantizer 604. Vector quantization information of the LPC coefficients is sent via line 605 to synthesis filter 606 and multiplexer 650.

A synthesis filter 606 outputs a synthesized signal corresponding to an excitation signal input through the line 628 using the vector quantization information of the LPC coefficients. The synthesized signal is output to subtractor 608 via line 607.

The subtractor 608 subtracts the synthesized signal input through the line 607 from the signal output from the preprocessing unit 602 input through the line 603 to generate the LPC residual signal. The LPC residual signal is transmitted via line 609 to the cognitive weight filter 610.

The cognitive weighting filter 610 allows the quantization noise to be less than or equal to the masking threshold in order to use the masking effect of the human auditory structure. Accordingly, the cognitive weight filter 610 outputs a signal including a weight to the pitch analyzer 612 so that the quantization noise of the LPC residual signal is minimized.

The pitch analyzer 612 searches for an open-loop pitch and a close-loop pitch with respect to the signal output from the cognitive weight filter 610. That is, the pitch analyzer 612 divides the signal output from the cognitive weight filter 610 into a plurality of pitch subframes, and analyzes and adapts the pitch of each subframe as in the standardized CLEP speech encoder. Output the codebook index and gain values.

The index of the adaptive codebook is transmitted to the multiplexer 650 through the line 614 while being transmitted to the pitch contribution remover 615 and the adaptive codebook 624 through the line 613. The gain value of the adaptive codebook is also provided to a gain value quantizer 629.

The pitch contribution remover 615 outputs a target signal required for fixed codebook search from the output signal of the cognitive weight filter 610 based on the index of the adaptive codebook. The pitch contribution remover 615 subtracts the pitch contribution y ₁ (n) from the line 611 to subtract the fixed codebook search target signal through the line 616 to the fixed codebook search unit 617 of the base layer 600. Will output The pitch contribution y ₁ (n) is obtained by equation (2).

The fixed codebook search unit 617 calculates a correlation d (n) between the target signal and the impulse response h (n) using the target signal x '(n) input through the line 611.

For example, assuming that the size of the subframe is 40 samples and the number of pulses in each layer is 4, the correlation d (n) may be defined as in Equation 1.

The fixed codebook search unit 617 searches a fixed codebook in the form of an algebraic codebook configured as in the example of Table 1 based on the impulse response h (n) and the correlation d (n). Referring to Table 1, the fixed codebook vector in the fixed codebook search unit 617 has a pulse size other than zero only at four positions. Accordingly, the correlation C corresponding to the magnitude of the correlation d (n) using the sign s of the pulse and the correlation d (n) may be defined as in Equation 2. The fixed codebook search unit 617 detects the correlation C by equation (2). The fixed codebook detector 617 detects the impulse response energy E by the equation (3).

The fixed codebook search unit 617 stores the correlation C and the energy E. The correlation degree C is stored by dividing the sign [d (i)] and its absolute value. sign [d (i)] is the sign of d (i). The energy E is stored in the form as shown in Equation 4. Equation 3 for energy E may be redefined as in Equation 5.

When the fixed codebook index and the gain value are obtained by the search, the fixed codebook search unit 617 transmits the fixed codebook index to the fixed codebook 619 and the multiplexer 650, and transmits the gain value to the gain value quantizer. To 629.

The fixed codebook 619 outputs a fixed codebook vector of the base layer 600 based on the index input through the line 618. The fixed codebook vector is constructed based on pulse position information m and sign information s. The fixed codebook vector output from the fixed codebook 619 is provided to the first multiplier 621 via line 620.

The first multiplier 621 multiplies the fixed codebook vector by the quantization gain value G _c for the gain value of the fixed codebook provided by gain value quantizer 629 and outputs the result via line 622. The signal output through the line 622 may be defined as the fixed codebook c _G (n) multiplied by the quantization gain value G _c of the fixed codebook vector of the base layer 600. The quantization gain value G _c is provided from a gain value quantizer 629.

When an adaptive codebook index is applied through line 613, the adaptive codebook 624 outputs an adaptive codebook vector corresponding to the adaptive codebook index. Via line 625 the adaptive codebook vector is provided to a second multiplier 626.

The second multiplier 626 multiplies the quantized gain value G _p for the gain value of the adaptive codebook by the adaptive codebook vector transmitted via the line 625 and outputs the result via line 627. The quantized gain value G _p is provided from a gain value quantizer 629.

The adder 623 adds the fixed codebook vector input through the line 622 and the adaptive codebook vector input through the line 627 to obtain an excitation signal. The excitation signal is output to the synthesis filter 606 via line 628.

The gain value quantizer 629 quantizes the gain value of the fixed codebook output from the fixed codebook search unit 617 and the gain value of the adaptive codebook output from the pitch analyzer 612, respectively. The gain value G _c obtained by quantizing the gain value of the fixed codebook is output to the first multiplier 621, and the gain value G _p obtained by quantizing the gain value of the adaptive codebook is output to the second multiplier 626. The quantized gain value G _c is also provided to a gain value quantizer 643 included in the sound quality enhancement layer 630.

The sound quality enhancement layer 630 is to provide additional bits in addition to the bits provided in the base layer 600 to improve sound quality restored as in the sound quality enhancement layer 130 of FIG. 1. 6 illustrates a configuration in which one voice enhancement layer 630 is connected to the base layer 600 for convenience of description, but a plurality of voice enhancement layers may be connected to the base layer 600.

The sound quality enhancement layer 630 includes a fixed codebook contribution calculator 631, a third adder 633, a synthesis filter 634, a cognitive weight filter 637, a fixed codebook search unit 639, a fixed codebook 641, A gain value difference quantizer 643 and a third multiplier 644.

When the fixed codebook c _G (n) obtained by multiplying the vector of the fixed codebook by the quantization gain value G _c from the first multiplier 621 of the base layer 600 is received, the fixed codebook contribution calculator 631 is expressed by Equation 15: The fixed codebook contribution y ₂ (n) is calculated.

In Equation 15, N is determined according to the number of samples constituting the size of the subframe. Therefore, as described in the pitch contribution removing unit 615, N is 40 when the size of the subframe is 40 samples. h (n) is the impulse response of the synthesis filter. The fixed codebook contribution calculated by the fixed codebook contribution calculator 631 is provided to the third adder 633 through the line 632.

The third adder 633 synthesizes via line 635 the fixed codebook contribution provided through line 632 in the target signal required for fixed codebook search of base layer 600 provided through line 616. A signal from which the synthesized signal provided from the filter 634 is removed is output.

가 승산된 고정 코드북이 입력되면, LPC 계수 추출 및 벡터 양자화기(604)에서 추출된 양자화된 LPC 계수를 사용하여 상기 입력되는 고정 코드북 신호를 합성한 신호를 출력한다. Synthesis filter 634 is a quantized fixed codebook gain value of the sound quality enhancement layer 630 quantized to a vector of fixed codebooks via line 647.

When a multiplying fixed codebook is input, a signal obtained by synthesizing the input fixed codebook signal is output using the quantized LPC coefficient extracted by the LPC coefficient extraction and the vector quantizer 604.

The cognitive weighting filter 637 performs cognitive weighting filtering on the signal input through the line 636 like the cognitive weighting filter 610, and outputs a target signal required for fixed codebook search in the sound quality enhancement layer 630. The target signal is transmitted through the line 638 to the fixed codebook search unit 639.

The fixed codebook search unit 639 searches the fixed codebook based on the input signal, such as the fixed codebook search unit 617 of the base layer 600, to obtain an index and a gain value of the fixed codebook. The index of the obtained fixed codebook is transmitted to the multiplexer 650 through the line 640 and to the fixed codebook 641. The gain value G _CE of the fixed codebook is transmitted to a gain value quantizer 643 via line 642.

The fixed codebook 641 outputs a fixed codebook vector of the sound quality enhancement layer 630 based on the input fixed codebook index. The fixed codebook vector can be constructed using the position information m of the pulse and the sign information s. The fixed codebook vector output from the fixed codebook 641 is provided to the third multiplier 644. The position of the pulse of the fixed codebook vector output from the fixed codebook 619 of the base layer 600 and the position of the pulse of the fixed codebook vector output from the fixed codebook 641 of the sound quality enhancement layer 630 may be the same.

를 출력한다. The gain difference quantizer 643 may include a gain value G _C obtained by quantizing a gain value of the fixed codebook output from the gain value quantizer 629 of the base layer 600 and a fixed codebook search unit of the sound quality enhancement layer 630 ( 639) using a log scale difference between the gain value G _CE of the unquantized fixed codebook output and the quantized fixed codebook gain value G _CE of improved quality layer 630 from the quantized gain value

Outputs

7 is a block diagram illustrating a preferred embodiment of a gain difference quantizer 643. The gain value quantizer 643 includes a first log scale converter 702, a second log scale converter 706, fourth and fifth multipliers 708 and 711, and a fourth adder 704.

When the quantized fixed codebook gain value G _C provided by the gain value quantizer 629 of the base layer 600 is input through the line 701, the first log scale converter 702 may use the fixed codebook gain. A log scale converted fixed codebook gain value corresponding to the value Gc is output through the line 703.

When the non-quantized gain value G _CE output from the fixed codebook search unit 639 of the sound quality enhancement layer 630 is input through the line 705, the log scale conversion is performed by the second load scale converter 706. The fixed codebook gain value is output via the line 707.

를 승산하고, 승산된 결과를 라인(708)을 통해 출력한다. The fourth multiplier 708 adjusts the gain difference adjustment value to the fixed codebook gain value input through the line 707.

Multiply by and output the multiplied result via line 708.

The fourth adder 704 outputs a difference value between the fixed codebook gain value input through the line 703 and the fixed codebook gain value input through the line 708 via the line 710.

The fifth multiplier 711 multiplies the scale difference element 10 by the input gain value difference to generate a log scale gain value difference G _DIFF 712.

The above-described operation of gain difference quantization 643 may be defined as in Equation 16 below.

는 로그 스케일 이득값간 차이값의 동적 범위가 최소가 되도록 하는 조정 값이다. 이득값 차 조정 값은 음성 부호화기의 종류에 따라 어떠한 값이 될 수도 있으며 실 예로 0.987이 사용된다. In Equation 16, G _c is a gain value of the fixed codebook quantized by the gain value quantizer 629, and G _CE is an unquantized gain value output from the fixed codebook search unit 639. In addition, the gain difference adjustment value

Is an adjustment value such that the dynamic range of the difference between logarithmic scale gain values is minimized. The gain difference adjustment value may be any value depending on the type of speech coder. For example, 0.987 is used.

를 출력한다. 상기 양자화된 이득값

는 라인(645)을 통해 제 3 승산기(644)로 출력되면서 라인(646)을 통해 다중화기(650)로 출력된다. The log scale gain difference 712 generated through the process as shown in Equation 16 is an analog signal and thus is quantized by a 3-bit scalar quantizer. Fixed codebook gain value of quantized sound quality enhancement layer 630 using the result quantized by 3-bit scalar quantizer

Outputs The quantized gain value

Is output to the multiplexer 650 via line 646 while being output to third multiplier 644 via line 645.

를 승산하고, 승산 결과를 라인(647)을 합성 필터(634)로 제공한다. The third multiplier 644 is a fixed codebook gain value of the quantized sound quality enhancement layer 6300 provided from the gain value difference quantizer 643 to a fixed codebook vector provided from the fixed codebook 641.

Multiply by < RTI ID = 0.0 > and < / RTI > provide a line 647 to the synthesis filter 634.

The multiplexer 650 is a fixed codebook index, a gain of the LPC coefficient quantization information, a fixed codebook index, an adaptive codebook index, a gain value quantization information provided from the base layer 600 and a sound quality enhancement layer provided from the sound quality enhancement layer 630. The value difference quantization information is output as a bit stream.

The bit streams of the base layer 600 and the sound quality enhancement layer 630 are transmitted separately. That is, as shown in FIG. 6, the bit stream of the sound quality enhancement layer 630 is transmitted after the bit stream of the base layer 600. Accordingly, the bit stream can be easily separated at the bit rate required for the decoding apparatus according to the network traffic conditions. For example, when the channel characteristic of the decoding apparatus is poor and only the bit stream of the base layer can be received, the decoding apparatus receives only the bit stream of the base layer from the bit stream transmitted by the bit rate extension speech encoding apparatus of FIG. 6. can do.

8 is a block diagram of a bit rate extended speech decoding apparatus according to another exemplary embodiment of the present invention. Referring to FIG. 8, the apparatus for decoding a bit rate extension speech includes a demultiplexer 802, an LPC coefficient decoder 803, a gain value decoder 804, a first fixed codebook decoder 805, and an adaptive codebook decoder ( 806, gain difference decoder 807, second fixed codebook decoder 808, multipliers 809, 810, 813, adders 811, 814, selection switch 812, synthesis filter 815 And a post-processing unit 816.

The bit rate extended speech decoding apparatus may selectively receive a bit stream transmitted from the bit rate extended speech encoding apparatus. That is, if only the bit stream for the base layer is received in the bit stream, the sound quality of the base layer can be restored, and if both the bit streams for the base layer and the sound quality enhancement layer are received, a more improved sound quality can be provided.

The demultiplexer 802 demultiplexes the received bit stream 801 with information of each element and outputs the demultiplexer. That is, the demultiplexer 802 converts the LPC coefficient quantization information to the LPC coefficient decoder 803, the gain value quantization information to the gain value decoder 804, and the gain difference quantization information to the gain value difference decoder 807. The fixed codebook index of the sound quality enhancement layer 630 is the second fixed codebook decoder 808, the fixed codebook index of the base layer 600 is the first fixed codebook decoder 805, and the adaptive codebook index is the adaptive codebook. Each is provided to the decoding unit 806.

The structure of the LPC coefficient decoder 803 is determined by the LPC coefficient extraction and vector quantizer 604 on the encoding apparatus side, and restores the LPC coefficients from the input LPC coefficient quantization information. The restored LPC coefficients are provided to the synthesis filter 815 and the post processor 816.

The structure of the gain value decoder 804 is determined by a gain value quantizer 629 on the encoding device side. The gain value decoder 804 decodes the input gain value quantization information. The gain value quantization information includes an adaptive codebook gain value and a fixed codebook gain value. Accordingly, the adaptive codebook gain value G _P and the fixed codebook gain value G _C in the base layer 600 are output from the gain value decoding unit 804, respectively.

The first fixed codebook decoder 805 decodes an input first fixed codebook index and outputs a first fixed codebook. The fixed codebook decoding method is determined by the search method in the fixed codebook search unit 617 of the encoding apparatus.

The adaptive codebook decoder 806 decodes an input adaptive codebook index and outputs an adaptive codebook.

The LPC coefficient decoder 803, the gain value decoder 804, the fixed codebook decoder 805, and the adaptive codebook decoder 806 are described in the base layer 600 transmitted from the demultiplexer 802. It may be defined as a decoding unit for decoding the encoding information.

The operation of the gain value difference decoder 807 and the second fixed codebook decoder 808 depends on the network traffic state or the processing capacity of the receiving terminal.

If it is determined that the gain value difference decoder 807 and the second fixed codebook decoder 808 are operated, the gain value difference decoder 807 decodes the input gain value difference quantization information. The second fixed codebook decoder 808 decodes the input second fixed codebook index. The gain value difference decoding method is determined by the gain value difference quantizer 643 on the encoding device side.

The decoding method of the second fixed codebook decoder 808 is determined by the second fixed codebook search unit 631 on the encoding apparatus side. The gain value difference decoder 807 and the second fixed codebook decoder 808 may be regarded as a decoding unit for decoding the encoding information in the sound quality enhancement layer 630 transmitted from the demultiplexer 902.

The multiplier 809 multiplies the fixed codebook gain value Gc of the base layer 600 restored by the gain value decoding unit 804 to the fixed codebook of the base layer output by the first fixed codebook decoding unit 805. Output a fixed codebook vector of the hierarchy.

를 제 2 고정 코드북 복호화부(808)에 의하여 출력된 음질 향상 계층의 고정 코드북에 승산하여 음질 향상 계층의 고정 코드북 벡터를 출력한다.The multiplier 810 is a fixed codebook gain value in the sound quality enhancement layer 630 reconstructed by the gain value difference decoder 807.

Is multiplied by the fixed codebook of the sound quality enhancement layer output by the second fixed codebook decoder 808 to output a fixed codebook vector of the sound quality enhancement layer.

The adder 811 adds the fixed codebook vector of the base layer output from the multiplier 809 and the fixed codebook vector of the sound quality enhancement layer output from the multiplier 810. Accordingly, the fixed codebook pulse in the decoding apparatus has an algebraic codebook pulse structure having multiple magnitudes by accumulating algebraic codebooks of the base layer and the sound quality enhancement layer. Accumulating the algebraic codebooks is to compensate for the disadvantages that occur in the existing fixed codebook structure in which all pulses of the fixed codebook have the same magnitude.

The selection switch 812 selectively transmits the signal output from the adder 811 and the fixed codebook vector of the base layer output from the multiplier 809. That is, when the decoding apparatus is not operated in the sound quality enhancement layer, the selection switch 812 selects and transmits the fixed codebook vector of the base layer output from the multiplier 809. When the encoder operates in the sound quality enhancement layer, the selection switch 812 transmits a signal output from the adder 811.

The multiplier 813 multiplies the decoded adaptive codebook output from the adaptive codebook decoding unit 806 by the gain value G _p of the adaptive codebook output from the gain value decoding unit 804 to output an adaptive codebook vector.

The adder 814 adds the fixed codebook vector selected by the selection switch 812 and the adaptive codebook vector output from the multiplier 813 to generate a reconstructed excitation signal.

The multiplier 810, the adder 811, and the selection switch 812 may decode the decoded signal in the unit for decoding the encoding information of the base layer and the unit for decoding the encoding information of the sound quality enhancement layer. It can be defined as a calculation unit that operates according to the operating environment.

The synthesis filter 815 synthesizes the excitation signal provided from the adder 814 using the reconstructed LPC provided from the LPC coefficient decoder 803 to reconstruct the speech signal.

The post processor 816 restores the voice signal transmitted from the synthesis filter 815. That is, the post processor 816 uses a LPC provided from the LPC coefficient decoder 803 to recover the speech signal, and then uses a high pass filter for filtering the signal output from the synthesis filter 815. Use

The synthesis filter 815 and the post processor 816 described above may be defined as a reconstruction unit for reconstructing a voice signal by synthesizing the signal output from the operation unit with the LPC output from the LPC coefficient decoder 803.

9 is a fixed codebook vector based on the position of the pulse searched by the fixed codebook search 901 of the base layer and the position of the pulse searched by the fixed codebook search 905 of the sound quality enhancement layer in the speech signal encoding apparatus of FIG. 8 is a view for explaining the magnitude of a pulse restored in the audio signal decoding apparatus of FIG.

9, a fixed codebook gain value G _c provided by the gain value decoding unit 804 is provided by a multiplier 809 to a fixed codebook vector 902 provided by the first fixed codebook decoding unit 805. A base layer fixed codebook vector 904 is generated that is multiplied and multiplied by a gain value.

The gain value G _CE provided by the gain value difference decoder 807 is multiplied by the multiplier 810 to the fixed codebook vector 906 provided by the second fixed codebook decoder 808 to multiply the gain value. A sound quality enhancement layer fixed codebook vector 908 is generated. The adder 811 generates a fixed codebook vector 910 obtained by adding the sound quality enhancement layer fixed codebook vector 908 and the base layer fixed codebook vector 904.

As can be seen based on the structure of the pulse generated in FIG. 9, the base layer fixed codebook vector 904 and the sound quality enhancement layer fixed codebook vector 908 are input to the adder 811 to add the final sound quality enhancement layer to which the two vectors are added. Generate a fixed codebook 910. Since the final sound quality enhancement layer fixed codebook 910 is formed by adding two fixed codebook vectors having different gain values, it is possible to form a fixed codebook having multiple sizes, thereby providing better sound quality.

10 is a flowchart illustrating an operation of a bit rate extended speech encoding method according to another exemplary embodiment of the present invention.

In operation 1001, the apparatus for encoding a speech signal preprocesses an input speech signal as in the preprocessing unit 602 of FIG. 6. In operation 1002, the apparatus for encoding a speech signal extracts LPC coefficients from a preprocessed speech signal and generates quantization information of the extracted LPC coefficients.

In operation 1003, the apparatus for encoding an audio signal detects a residual signal of the LPC coefficients through the synthesis filter 606 in the preprocessed signal. In operation 1004, the apparatus for encoding a speech signal filters the detected residual signal as in the cognitive weighting filter 610 of FIG. 6 to output a cognitive weighted signal.

In operation 1005, the apparatus for encoding a speech signal analyzes the pitch of the cognitive weighted signal as in the pitch analyzer 612 of FIG. 6, and calculates the pitch contribution of the cognitive weighted signal using the analyzed result. The removal unit 615 removes the adaptive codebook gain value and the adaptive codebook index.

In operation 1006, the apparatus for encoding a speech signal searches for a base layer fixed codebook as in the fixed codebook search unit 617 of the base layer 600 of FIG. 6 to generate a fixed codebook gain value and a fixed codebook index.

In operation 1007, the apparatus for encoding a speech signal quantizes the detected fixed codebook gain value and the detected adaptive codebook gain value as in the gain value quantizer 629 of FIG. 6.

In operation 1008, the apparatus for encoding a speech signal may extract the excitation signals of the fixed codebook vector and the adaptive codebook vector generated in the base layer 600 using the vector quantized LPC coefficients from the synthesis filter 606 of FIG. 6. Synthesize together.

In operation 1009, the apparatus for encoding a speech signal removes the influence of the target signal for the fixed codebook search in the base layer 600 and the previous LPC synthesis signal of the sound quality enhancement layer 630. Generate a target signal for fixed codebook search as in < RTI ID = 0.0 > That is, a signal obtained by removing the fixed codebook contribution of the base layer and the previous LPC synthesis signal detected by the sound quality enhancement layer 630 from the target signal detected by the base layer 600 is used as the target signal in the sound quality enhancement layer.

In operation 1010, the apparatus for encoding a speech signal performs fixed codebook search of the sound quality enhancement layer 630 using the target signal detected in operation 1009 to obtain a fixed codebook gain value of the sound quality enhancement layer and a fixed codebook index of the sound quality enhancement layer. Create each.

In operation 1011, the apparatus for encoding a speech signal quantizes a log scale difference between a gain value of the quantized fixed codebook of the base layer and an unquantized fixed codebook gain value of the sound quality enhancement layer. The fixed codebook search and gain value quantization processes in the above-described sound quality enhancement layer may be performed a plurality of times as a plurality of sound quality enhancement layers are provided. When the sound quality enhancement layer processing is performed a plurality of times, the quality of the speech signal restored by that can be improved.

In operation 1012, the apparatus for encoding a speech signal passes the fixed codebook vector (or excitation signal) generated in the sound quality enhancement layer and passes through the synthesis filter 634 of FIG. 6 to output the synthesized signal.

In operation 1013, the apparatus for encoding a speech signal includes linear prediction coefficient quantization information obtained through the above-described steps, a fixed codebook index of a base layer, an adaptive codebook index of a base layer, a gain value of a fixed codebook of a base layer, and an adaptive codebook of a base layer. The gain value, the fixed codebook index of the sound quality enhancement layer, and the gain value quantization information are multiplexed in the form of a bit stream and sent to the voice signal decoding apparatus.

11 is a flowchart illustrating an operation of a bit rate extended speech decoding method according to another exemplary embodiment of the present invention.

In operation 1101, the apparatus for decoding a speech signal demultiplexes the received bit stream into information of each component, such as the demultiplexer 802 of FIG. 8.

In operation 1102, the voice signal decoding apparatus decodes the demultiplexed signal. That is, the LPC coefficient decoder 803, the gain value decoder 804, the first fixed codebook decoder 805, the adaptive codebook decoder 806, the gain value difference decoder 807, and the second of FIG. Like the fixed codebook decoder 808, the demultiplexed signal is decoded.

In operation 1103, the voice signal decoding apparatus selectively transmits the fixed codebook of the sound quality enhancement layer and the fixed codebook of the base layer according to the operating conditions of the voice signal decoding apparatus, and optionally the gain value. That is, when the speech signal decoding apparatus is operated in the sound quality enhancement layer, the gain value of the fixed codebook of the base layer is added to the fixed codebook of the sound quality enhancement layer and the base layer fixed codebook multiplied by the gain value of the fixed codebook of the restored sound quality enhancement layer. The multiplied fixed codebook is added and transmitted. On the other hand, if the speech signal encoding apparatus does not operate in the sound quality enhancement layer, the fixed codebook multiplying the gain value of the fixed codebook of the base layer is transmitted to the decoded fixed codebook of the base layer.

In operation 1104, the apparatus for decoding speech signals synthesizes a codebook selectively transmitted in operation 1103 using the LPC coefficients decoded in operation 1102.

In operation 1105, the voice signal decoding apparatus post-processes the post-processing unit 816 to generate a reconstructed voice signal.

So far I looked at the center of the preferred embodiment for the present invention. Those skilled in the art will appreciate that the present invention can be implemented in a modified form without departing from the essential features of the present invention. Therefore, the disclosed embodiments should be considered in descriptive sense only and not for purposes of limitation. The scope of the present invention is shown in the claims rather than the foregoing description, and all differences within the scope will be construed as being included in the present invention.

According to the present invention described above, according to the present invention described above, by presenting a structure capable of extending the bit rate without changing the existing standardized CLEP speech coding structure and the system having a conventional standardized CLEP speech coding apparatus and Compatible

In addition, according to the embodiment of the present invention described above, by fixing the fixed codebook search target signal of the base layer and the fixed codebook search target signal of the sound quality enhancement layer, the codebook searched in the sound quality enhancement layer is not stored for the next frame. It does not affect the behavior of the base layer.

When the fixed codebook search of the sound quality enhancement layer is used, the parameter value obtained when the fixed codebook search of the base layer is used may reduce the amount of computation required for the fixed codebook search.

In addition, according to another embodiment of the present invention, the target signal required for the fixed codebook search of the sound quality enhancement layer is a fixed codebook target signal of the base layer through the fixed codebook contribution of the base layer and the synthesis filter of the sound quality enhancement layer By removing the synthesized signal of the fixed codebook of the previous sound quality enhancement layer provided, a more accurate fixed codebook search can be expected as the fixed codebook search using the target signal dedicated to the sound quality enhancement layer is performed.

Moreover, the position of the searched pulse in the sound quality enhancement layer and the position of the searched pulse in the base layer can be the same, overcoming the limitation that the magnitude of each pulse in the algebraic codebook has the same magnitude, and the pulses of the final fixed codebook are Since it can improve the sound quality of the restored voice signal.

The gain value of the sound quality enhancement layer quantizes a difference between the quantized gain value of the base layer and the gain value of the sound quality enhancement layer, and transmits a quantized value of a gain difference having a relatively small dynamic range, thereby obtaining a gain in the sound quality enhancement layer. The bits required for value quantization can be saved.

Claims (34)

delete delete delete delete delete delete In the audio signal encoding apparatus, A base layer for linear prediction encoding filtering the input speech signal and generating an excitation signal corresponding to the filtered speech signal by fixed codebook searching and adaptive codebook searching; And A fixed codebook search unit for searching a fixed codebook using a parameter generated according to the fixed codebook search in the base layer; A gain value quantizer for detecting a difference between a first fixed codebook gain value generated by the fixed codebook search of the base layer and a second fixed codebook gain value output from the fixed codebook search unit, and quantizing the detected difference; It includes a plurality of sound quality enhancement layer, And a multiplexer for multiplexing the signal generated in the base layer and the signal generated in the sound quality enhancement layer. An audio signal decoding apparatus for decoding an encoded speech signal divided into a base layer and at least one sound quality enhancement layer, A first decoding unit for decoding encoded information in a base layer among the encoded speech signals; A second decoding unit for restoring encoding information in a sound quality enhancement layer among the encoded speech signals according to an operating environment of the speech signal decoding apparatus; A calculation unit configured to calculate a signal restored in the first decoding unit and a signal restored in the second decoding unit according to an operating environment of the speech signal decoding apparatus; And a speech signal reconstruction unit for reconstructing the speech signal by synthesizing the signal output from the calculation unit by using the linear prediction coding coefficients output from the first decoding unit. The method of claim 8, wherein the first decoding unit, A linear prediction coding coefficient decoder for decoding the linear prediction coding coefficient quantization information included in the encoding information of the base layer; A first fixed codebook decoder which decodes a fixed codebook index included in the encoding information in the base layer; An adaptive codebook decoder which decodes an adaptive codebook index included in the encoding information in the base layer; And a gain value decoder which decodes the fixed codebook gain value and the adaptive codebook gain value included in the encoding information in the base layer, respectively. The method of claim 9, wherein the second decoding unit, A gain value difference decoder for decoding quantization information of the difference between the fixed codebook gain values included in the encoded information in the speech enhancement layer; And a second fixed codebook decoder for decoding the fixed codebook index included in the encoding information in the sound quality enhancement layer. The method of claim 10, wherein the calculation unit, A first adder for adding the decoded fixed codebook gain value output from the gain value decoder and the decoded gain value difference output from the gain value difference decoder; A first selection switch for transmitting the decoded fixed codebook gain value output from the gain value decoding unit or the gain value output from the first adder according to an operating condition of the speech signal decoding apparatus; A second adder for adding a fixed codebook of a decoded sound quality enhancement layer output from the second fixed codebook decoder and a fixed codebook of a decoded base layer output from the first fixed codebook decoder; A second selection switch for transmitting a signal output from the second adder or the decoded fixed codebook output from the first fixed codebook decoder according to an operating condition of the voice signal decoding apparatus; A first multiplier for multiplying a gain value of the decoded adaptive codebook output from the adaptive codebook decoder and the decoded adaptive codebook output from the gain value decoder; A second multiplier that multiplies the signal output from the first selection switch with the signal output from the second selection switch; And a third adder configured to add a signal output from the first multiplier and a signal output from the second multiplier. The method of claim 11, wherein the audio signal recovery unit, A synthesis filter for synthesizing a signal output from the third adder using the linear prediction coding coefficients; And a post-processing unit for obtaining the reconstructed speech signal using the linear prediction coefficients and the signal output from the synthesis filter. The method of claim 8, wherein the second decoding unit, A gain value difference decoder for decoding quantization information of the difference between the fixed codebook gain values included in the encoded information in the speech enhancement layer; And a fixed codebook decoder for decoding the fixed codebook index included in the encoding information in the sound quality enhancement layer. The method of claim 9, wherein the second decoding unit, A gain value difference decoding unit for decoding the quantization information of the difference between the fixed codebook logscale gain values included in the encoding information in the sound quality enhancement layer; And a second fixed codebook decoder for decoding the fixed codebook index included in the encoding information in the sound quality enhancement layer. The method of claim 14, wherein the calculation unit, A first adder for adding a fixed codebook of a decoded sound quality enhancement layer output from the second fixed codebook decoder and a fixed codebook of a decoded base layer output from the first fixed codebook decoder; A selection switch for selectively transmitting a signal output from the first adder or a fixed codebook of the decoded base layer output from the first fixed codebook decoder according to an operating condition of the speech signal decoding apparatus; And a second adder for adding the signal output from the selection switch and the adaptive codebook of the decoded base layer output from the adaptive codebook decoder. The method of claim 15, wherein the audio signal recovery unit, A synthesis filter for synthesizing a signal output from the second adder using the linear prediction coding coefficients; And a post-processing unit for obtaining the reconstructed speech signal using the linear prediction coefficients and the signal output from the synthesis filter. The method of claim 8, wherein the second decoding unit, A gain value difference decoder which decodes quantization information of the difference between the fixed codebook logscale gain values included in the encoded information in the speech enhancement layer; And a fixed codebook decoder for decoding the fixed codebook index included in the encoding information in the sound quality enhancement layer. delete delete delete delete A speech signal decoding method for decoding a speech signal encoded with a base layer and at least one sound quality enhancement layer, Decoding the encoded speech signal; Selectively transmitting the codebook for the base layer decoded in the decoding step and the codebook for the sound quality enhancement layer according to the operating condition of the speech signal decoding; And generating a reconstructed speech signal by combining the selectively transmitted codebook and the linear prediction coefficients decoded in the decoding step. The method of claim 22, wherein the decoding step, And demultiplexing the encoded speech signal into encoding information of a base layer and encoding information of a sound quality enhancement layer, and decoding the demultiplexed encoding information. The method of claim 23, wherein the voice signal decoding method, Restoring the gain value of the fixed codebook in the sound quality enhancement layer by adding a difference between the gain value of the fixed codebook in the decoded base layer and the gain value of the fixed codebook included in the decoded sound quality enhancement layer in the decoding step. Voice signal decoding method further comprising. In an audio signal encoding apparatus. A base layer for filtering an input speech signal using linear predictive coding and generating an excitation signal of the filtered speech signal by fixed codebook searching and adaptive codebook searching; At least one sound quality enhancement layer for searching the fixed codebook by using a signal obtained by removing the contribution of the fixed codebook of the base layer from the fixed codebook search target signal of the base layer, And a multiplexer for multiplexing the signal generated in the base layer and the signal generated in the sound quality enhancement layer and outputting the multiplexed signal. The fixed codebook contribution y ₂ (n) of the base layer is detected by calculating the impulse response of the synthesis code and the fixed codebook c _G multiplied by the quantization gain value of the fixed codebook of the base layer as shown in the following equation. Voice signal encoding apparatus, characterized in that.

The audio signal encoding of claim 25, wherein the sound quality enhancement layer further removes a signal obtained by synthesizing a fixed codebook signal generated in the sound quality enhancement layer using the linear prediction coding coefficients from a target signal of the base layer. Device. 26. The method of claim 25, wherein, in the fixed codebook search of the sound quality enhancement layer, the log scale value of the first gain value obtained by the fixed codebook search of the base layer and the second gain obtained by the fixed codebook search in the sound quality enhancement layer. And a function for multiplying the quantized gain value of the sound quality enhancement layer obtained by quantizing the difference between logarithmic scale values and the fixed codebook vector obtained by the fixed codebook search in the sound quality enhancement layer. . The method of claim 25, wherein if there are a plurality of sound quality enhancement layers, And the multiplexer multiplexes quantized information on a difference between a fixed codebook index output from a plurality of sound quality enhancement layers and a log scale gain value of the fixed codebook. The apparatus of claim 25 or 27, wherein the sound quality enhancement layer performs the fixed codebook search after the cognitive weighted filtering of the target signal. In the audio signal encoding apparatus, A base layer for linearly predicting and encoding the input speech signal and generating an excitation signal corresponding to the filtered speech signal by fixed codebook search and adaptive codebook search; A search unit for searching for a fixed codebook by using a signal from which the fixed codebook contribution of the base layer is removed from the fixed codebook search target signal of the base layer as a fixed codebook search target signal of a sound quality enhancement layer; Detecting a difference between a log scale gain value of a first fixed codebook generated by the fixed codebook search of the base layer and a log scale gain value of a second fixed codebook output from the fixed codebook search unit, and quantizing the detected difference A plurality of sound quality enhancement layers including a log scale gain value difference quantizer, A multiplexer for multiplexing the signal generated in the base layer and the signal generated in the sound quality enhancement layer, And the sound quality enhancement layer further removes a signal obtained by synthesizing the fixed codebook using the linear predictive coding coefficients from the fixed codebook search target signal of the sound quality enhancement layer. The voice signal coding method is A base layer processing step of extracting a linear prediction coefficient of the input speech signal and generating an excitation signal corresponding to the input speech signal by fixed codebook searching and adaptive codebook searching; A sound quality enhancement layer processing step of searching for a fixed codebook by using a signal from which the fixed codebook contribution of the base layer is removed from the fixed codebook search target signal of the base layer as a fixed codebook search target signal of a sound quality enhancement layer; And multiplexing signals generated by the base layer processing step and the sound quality enhancement layer processing step. 33. The method of claim 32, wherein the fixed codebook search target signal of the sound quality enhancement layer further removes a signal obtained by synthesizing the fixed codebook using linear prediction coding coefficients in the sound quality enhancement layer from the target signal of the base layer. Speech signal coding method. 33. The method of claim 32, wherein the sound quality enhancement layer processing step includes a load scale gain value of a fixed codebook obtained by the fixed codebook search in the base layer processing step and a log scale of a gain value obtained by the fixed codebook search of the sound quality enhancement layer. And quantizing the difference between the gain values.

Images